Cochlear implants (inner-ear prostheses) are a possibility to help profoundly deaf or severely hearing impaired persons. Unlike conventional hearing aids, which just apply an amplified and modified sound signal, a cochlear implant is based on direct electrical stimulation of the acoustic nerve. The intention of a cochlear implant is to electrically stimulate neural structures in the inner ear in such a way that a hearing sensation most similar to normal hearing is obtained.
FIG. 1 shows a conventional cochlear prosthesis. The cochlear prosthesis essentially consists of two parts, the speech processor 101 that is typically positioned externally proximate the ear, and the implanted stimulator 105. The speech processor 101 includes the power supply (batteries) of the overall system and is used to perform signal processing of the acoustic signal to extract the stimulation parameters. The stimulator 105 generates the stimulation patterns and conducts them to the nerve tissue by means of an electrode array 107 that extends into the scala tympani 109 in the inner ear. The connection between speech processor and stimulator is established either by means of a radio frequency link (transcutaneous) using primary coils 103 and secondary coils within stimulator 105, or by means of a plug in the skin (percutaneous).
One successful stimulation strategy is the so called “continuous-interleaved-sampling strategy” (CIS), as described by Wilson B. S., Finley C. C., Lawson D. T., Wolford R. D., Eddington D. K., Rabinowitz W. M., “Better speech recognition with cochlear implants,” Nature, vol. 352, 236-238 (July 1991) [hereinafter Wilson et al., 1991], which is incorporated herein by reference. Signal processing for CIS in the speech processor involves the following steps:                a. splitting up of the audio frequency range into spectral bands by means of a filter bank,        b. envelope detection of each filter output signal, and        c. instantaneous nonlinear compression of the envelope signal (map law).        
According to the tonotopic organization of the cochlea, each stimulation electrode in the scala tympani is associated with a band pass filter of the external filter bank. For stimulation, symmetrical biphasic current pulses are applied. The amplitudes of the stimulation pulses are directly obtained from the compressed envelope signals (step (c) of above). These signals are sampled sequentially, and the stimulation pulses are applied in a strictly non-overlapping sequence. Thus, as a typical CIS-feature, only one stimulation channel is active at one time. The overall stimulation rate is comparatively high. For example, assuming an overall stimulation rate of 18 kpps, and using a 12-channel filter bank, the stimulation rate per channel is 1.5 kpps. Such a stimulation rate per channel usually is sufficient for adequate temporal representation of the envelope signal.
The maximum overall stimulation rate is limited by the minimum phase duration per pulse. The phase duration cannot be chosen arbitrarily short, because the shorter the pulses, the higher the current amplitudes have to be to elicit action potentials in neurons, and current amplitudes are limited for various practical reasons. For an overall stimulation rate of 18 kpps, the phase duration is 27 μs, which is at the lower limit.
Each output of the CIS band pass filters can roughly be regarded as a sinusoid at the center frequency of the band pass filter, which is modulated by the envelope signal. This is due to the quality factor Q=3 of the filters. In case of a voiced speech segment, this envelope is approximately periodic, and the repetition rate is equal to the pitch frequency.
In the current CIS-strategy, the envelope signals only are used for further processing, i.e., they contain the entire stimulation information. For each channel, the envelope is represented as a sequence of biphasic pulses at constant repetition rate. As a characteristic feature of CIS, this repetition rate (typically 1.5 kpps) is equal for all channels, and there is no relation to the center frequencies of the individual channels. It is intended that the repetition rate is not a temporal cue for the patient, i.e., it should be sufficiently high, so that the patient does not perceive tones with a frequency equal to the repetition rate. The repetition rate is usually set to more than twice the bandwidth of the envelope signals (Nyquist theorem).
Electrode Configuration of a 12-Channel Cochlear Implant Using Monopolar Stimulation
FIG. 2 shows an example of an electrode configuration used in a 12-channel cochlear implant as described in U.S. Pat. No. 6,600,955. An electrode array containing 12 electrode contacts 201 (black dots) is positioned within the scala tympani of the cochlea. Each of these electrodes 201 is connected to a capacitor C 203 and a pair of current sources 205 and 207, whereby the second ports of current sources 205 and 207 are connected to implant ground GND 209 and implant supply voltage VCC 211, respectively. Current sources 205 and 207 may be implemented, for example, using P-channel and N-channel MOS field effect transistors, respectively. Thus, for convenience, the sources 205 and 207 are designated as P-sources and N-sources. Reference electrode 213 is positioned outside the cochlea and connected to a pair of switches 215 and 217, whereby the second ports of switches 215 and 217 are connected to implant ground GND and implant supply voltage VCC, respectively.
A simplified lumped-element model of this configuration is shown in FIG. 3. Impedances ZI 301 represent the interface impedances between the metal surfaces of the intra-cochlear electrode contacts and the fluid within the scala tympani. Impedance ZI,REF 303 represents the interface impedance of the reference electrode. The intra-cochlear fluid is represented by the ohmic resistors RS 305. Since the cross-sectional area is changing along the scala tympani, usually a variable RS is assumed, as described in Kral A., Hartmann R., Mortazavi D., and Klinke R., “Spatial Resolution of Cochlear Implants: The Electrical Field and Excitation of Auditory Afferents,” Hearing Research 121, pp. 11-28, 1998, which is hereby incorporated herein by reference. Resistors RB 307 describe the bony structures in which the cochlea is embedded, and they are also position-dependent. The spatial dependencies are of minor importance and therefore, for convenience, Rs and RB are assumed to be constant. Besides, an infinite ladder network Rs/RB is assumed. The stimulation current passes the highly resistive structures on its way to the reference electrode.
Impedances ZI and ZI,REF in general are complex and frequency-dependent. However, in-vitro measurements of the impedances show that for the electrode geometries and the very short pulsatile stimulation waveforms used in cochlear implant applications, the interface impedances can be assumed to be purely ohmic.
As described in U.S. Pat. No. 6,600,955, a stimulation configuration as shown in FIG. 3 may be used to generate either (a) single non-simultaneous stimulation pulses, or (b) simultaneous pulses which are “sign-correlated”. For example, the two phases of a single symmetric, biphasic pulse in one electrode are produced by first activating one of the P-sources 313 associated to this electrode and closing switch 315, and then activating the associated N-source 311 and closing switch 317. In the first phase of this pulse, the current is flowing from the pair of associated current sources via the ladder network to the pair of switches, and in the second phase the current direction is reversed. If current amplitudes and phase durations of the two phases are equal, the pulse is charge balanced, that is, no net charge is delivered to the ladder network.
If more than one stimulation pulses are applied simultaneously, such pulses are subject to “sign-correlation”, i.e., either several P-sources are activated simultaneously and switch 315 is closed, or several N-sources are activated simultaneously and switch 317 is closed, but no mixture between activated P- and N-sources occurs. This ensures that the sum of currents is always flowing through the reference electrode (i.e., impedance ZI,REF). Such a stimulation arrangement is designated as “distributed monopolar”.
The electrical potentials which occur, for example, during the first phase of a single biphasic pulse are explained with the help of FIG. 4. Let P-source 401 produce a particular amplitude IP causing a voltage drop UP (note that the associated N-source 403 is inactive in this phase). Assuming capacitor 405 as being uncharged prior to the pulse, current IP will cause a voltage UC across capacitor 405, which is linearly increasing with time. However, assuming a sufficiently high capacitance, only a comparatively small voltage will drop across capacitor 405 the end of the first pulse phase. Typically, UC is not larger than some tens of millivolts, and thus is usually is negligible as compared to other voltage drops in the ohmic network. Interface impedance ZI causes a considerable voltage drop UI=ZIIP. Current IP is distributed within the infinite ladder network composed of horizontal resistors RS and vertical resistors RB. The distribution of voltage drops across vertical resistors RB will show exponential behavior, where the maximum voltage drop UB occurs in resistor 409, and the voltage drops across the neighboring resistors RB at both sides will decay exponentially, i.e., αUB in resistors 411 and 413, α2UB in resistors 415 and 417, α3UB in resistors 419 and 421, etc. Factor α is a function of ratio RS/RB only, and a short calculation yields
  α  =      1    +                  R        S                    2        ⁢                  R          B                      -                                                      R              S                                      R              B                                +                                    (                                                R                  S                                                  2                  ⁢                                      R                    B                                                              )                        2                              .      
The sum of all currents flowing through resistors RB is again IP, which is flowing back to implant ground via impedance ZI,REF 423 and the closed switch 425. Voltage UI,REF across ZI,REF is given by UI,REF=ZI,REFIP, and assuming ideal switches, there is no voltage drop across the closed switch 425. Summing up all voltage drops yields the implant supply voltage VCC, that is,VCC=UP+UC+UI+UB+UI,REF.  (1)
The overall power consumption of such a circuit isPTOT=VCCIP  (2)
In the present application, PTOT is preferably as small as possible. For a given current amplitude IP, the overall power consumption is minimized, if the implant supply voltage is minimized.
As a typical numeric example, assume interfaces impedances ZI=5 kΩ and ZI,REF=250Ω, ladder network impedances RS=450Ω and RB=9 kΩ (resulting in α=0.8), and a current amplitude IP=800 μA. These assumptions yield UI=4V, UB=0.8V, and UI,REF=0.2V. Inserting in Eq. (1) and neglecting voltage UC across the capacitor yields VCC−UP=UI+UB+UI,REF=5V. Assuming that the P-source 401 can be operated with negligible voltage UP yields a minimum implant supply voltage VCC=5V. Inserting in Eq. (2) yields overall power PTOT=4 mW. Obviously, 80% of PTOT is absorbed by interface impedance ZI, i.e., PI=UIIP=3.2 mW, and this power does not contribute to the stimulation itself. Thus, any reduction of voltage drop UI is desirable with respect to both the reduction of the implant supply voltage and the reduction of the stimulation power consumption.
One approach for reducing the voltage drop across ZI is to try to reduce ZI itself. For example, using larger electrode surfaces would reduce ZI. However, the size of the electrode surfaces typically cannot be increased further, because geometrical limits such as electrode distances have already been reached. Another approach is based on the observation that ZI is not stable over time, but increasing in the weeks after the implantation. The reasoning is that the growth of a particular tissue covers the electrode surfaces. Giving corticoids during surgery seems to reduce this additional tissue growth and keep the impedance at least at its initial value.